“…CD scores were calculated as the number of instances of CD divided by the number of words spoken to account for verbosity (as recommended by previous researchers; Hirsch & Leff, 1971; Miklowitz & Stackman, 1992). This coding protocol has been shown to have good reliability and construct validity (Velligan et al, 1990), and has been previously used with clinical (Velligan, Funderburg, Giesecke, & Alexander, 1995; Velligan et al, 1996) and high-risk populations (Velligan et al, 1988). The system has also been previously applied to 5-min speech samples (Kymalainen, 2005; Kymalainen, Weisman, Rosales, & Armesto, 2006), and to natural speech samples (Docherty, 1993).…”